Created attachment 116018 [details] dmesg log ==System Environment== -------------------------- Regression: Yes, don't see this issue on drm-intel-testing-2015-05-08. Non-working platforms: BSW BIOS: V69 ==kernel== -------------------------- -testing: drm-intel-testing-2015-05-22 (fails) ==Bug detailed description== ----------------------------- "testdisplay -d 32" core dump on BSW: root@x-bsw14:/GFX/Test/Intel_gpu_tools/intel-gpu-tools/tests# ./testdisplay -d 32 using depth 32 CRTC(20):[0] 1920x1080 60 1920 1966 1996 2080 1080 1082 1086 1112 0xa 0x48 138780 failed to set mode (1920x1080@60Hz): Invalid argument Received signal SIGSEGV. Segmentation fault (core dumped) "testdisplay -d 16" and "testdisplay -d 24" are fine. ==Reproduce steps== ---------------------------- 1. Reboot and login into BSW 2. ./testdisplay -d 32
root@xfzheng:~/git/drm-intel# git bisect start root@xfzheng:~/git/drm-intel# git bisect good c457879f0d8f664050e46092afde46fec58c9e67 root@xfzheng:~/git/drm-intel# git bisect bad 5d96d8afcfbb1500272756747e0bfcdbebf3b070 Bisecting: a merge base must be tested [214a2b7fab215b1e979fbae51225b01b8fc58288] drm/i915: Update DRIVER_DATE to 20150508 root@xfzheng:~/git/drm-intel# git bisect good Bisecting: 404 revisions left to test after this (roughly 9 steps) [6c9d370c16aad97cfc6de68666634eaabac2b048] Merge branch 'upstream' of git://git.linux-mips.org/pub/scm/ralf/upstream-linus root@xfzheng:~/git/drm-intel# git bisect good Bisecting: 201 revisions left to test after this (roughly 8 steps) [518af3cb8ccaf32057db6046e241ec393d6c7b98] Merge branch 'master' of git://git.linux-mips.org/pub/scm/ralf/upstream-linus root@xfzheng:~/git/drm-intel# git bisect good Bisecting: 117 revisions left to test after this (roughly 7 steps) [e26081808edadfd257c6c9d81014e3b25e9a6118] Linux 4.1-rc4 root@xfzheng:~/git/drm-intel# git bisect good Bisecting: 58 revisions left to test after this (roughly 6 steps) [d94b5030d26b4f45510a092262bc2b542a00bd7c] drm/i915: Ensure cache flushes prior to doing CS flips root@xfzheng:~/git/drm-intel# git bisect good Bisecting: 29 revisions left to test after this (roughly 5 steps) [8504c74c7ae48b4b8ed1f1c0acf67482a7f45c93] drm/i915: Preserve ddi_pll_sel when allocating new pipe_config root@xfzheng:~/git/drm-intel# git bisect good Bisecting: 14 revisions left to test after this (roughly 4 steps) [1e8df16778b0d8fd8102b3ee799b028f8f961089] drm/i915/skl: Swapping 90 and 270 to be compliant with Xrandr root@xfzheng:~/git/drm-intel# git bisect bad Bisecting: 7 revisions left to test after this (roughly 3 steps) [0c9b371550f40f8273a835f1e1055445cc3e2498] drm/i915: add HAS_DP_MST feature test macro root@xfzheng:~/git/drm-intel# git bisect good Bisecting: 3 revisions left to test after this (roughly 2 steps) [6c0fd451bdde9a6d0ac408bd078351890228e465] drm/i915: Tighten the exposure ARGB/ABGR 8888 formats root@xfzheng:~/git/drm-intel# git bisect bad Bisecting: 1 revision left to test after this (roughly 1 step) [e0681e3841118b470b312b7c62c90af9364fce8d] drm/i915/bxt: Move around lane stagger calculation root@xfzheng:~/git/drm-intel# git bisect good Bisecting: 0 revisions left to test after this (roughly 0 steps) [c965d99557fd48e8c57d793ec22458e508cbf676] drm/i915: Fix 'suspedn' typo root@xfzheng:~/git/drm-intel# git bisect good 6c0fd451bdde9a6d0ac408bd078351890228e465 is the first bad commit commit 6c0fd451bdde9a6d0ac408bd078351890228e465 Author: Damien Lespiau <damien.lespiau@intel.com> Date: Tue May 19 12:29:16 2015 +0100 drm/i915: Tighten the exposure ARGB/ABGR 8888 formats ARGB8888 is used for cursors on all platforms so we need to allow it everywhere. ABGR8888 is currently only honoured: - on VLV/CHV in sprite planes - on SKL+ for primary and sprite planes so only allow it for those platforms. Note that we only support ARGB8888/ABGR8888 on the primary plane for SKL/BXT because we have in line of sight the pipe bottom color on those platforms and because the primary plane programming on VLV/CHV doesn't anything different for those formats today. v2: Fix the logic to forbid the creation ABGR2101010 fbs (Ville) v3: Still allow the creation of ARGB8888 fbs now that cursor planes use real fb objects (found by PRTS). Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Damien Lespiau <damien.lespiau@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> :040000 040000 322b628bb8364c3ff3f13826349abe6c3b49eff7 772ec232149432c26c2a7c92055c623c2bb13d39 M drivers
Can you provide a gdb backtrace for testdisplay?
(gdb) run -d 32 Starting program: /GFX/Test/Intel_gpu_tools/intel-gpu-tools/tests/testdisplay -d 32 [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1". using depth 32 CRTC(20):[0] 1920x1080 60 1920 1966 1996 2080 1080 1082 1086 1112 0xa 0x48 138780 failed to set mode (1920x1080@60Hz): Invalid argument Program received signal SIGSEGV, Segmentation fault. INT_cairo_surface_destroy (surface=0x4088980000000000) at cairo-surface.c:950 950 cairo-surface.c: No such file or directory. (gdb) bt #0 INT_cairo_surface_destroy (surface=0x4088980000000000) at cairo-surface.c:950 #1 0x000000000040f574 in igt_remove_fb (fd=4, fb=fb@entry=0x7fffffffdd90) at igt_fb.c:866 #2 0x00000000004065eb in set_mode (c=0x639d80) at testdisplay.c:394 #3 update_display () at testdisplay.c:499 #4 0x0000000000405544 in main (argc=<optimized out>, argv=<optimized out>) at testdisplay.c:772
Created attachment 116069 [details] [review] Fix for igt Please tests the attach patch for i-g-t.
(In reply to Ander Conselvan de Oliveira from comment #4) > Created attachment 116069 [details] [review] [review] > Fix for igt > > Please tests the attach patch for i-g-t. Yes, testdisplay will not core dump with the patch, but it still complains "invalid argument": root@x-bsw14:/GFX/Test/Intel_gpu_tools/intel-gpu-tools/tests# ./testdisplay -d 32 using depth 32 CRTC(20):[0] 1920x1080 60 1920 1966 1996 2080 1080 1082 1086 1112 0xa 0x48 138780 failed to set mode (1920x1080@60Hz): Invalid argument And is this a kernel bug or igt bug?
(In reply to Jeff Zheng from comment #5) > (In reply to Ander Conselvan de Oliveira from comment #4) > > Created attachment 116069 [details] [review] [review] [review] > > Fix for igt > > > > Please tests the attach patch for i-g-t. > > Yes, testdisplay will not core dump with the patch, but it still complains > "invalid argument": > root@x-bsw14:/GFX/Test/Intel_gpu_tools/intel-gpu-tools/tests# ./testdisplay > -d 32 > using depth 32 > CRTC(20):[0] 1920x1080 60 1920 1966 1996 2080 1080 1082 1086 1112 0xa 0x48 > 138780 > failed to set mode (1920x1080@60Hz): Invalid argument > > > And is this a kernel bug or igt bug? Well, the test indicates Damien's patch work as intended. Using an ARGB8888 fb for the primary plane in platforms other than SKL+ shouldn't work. The crash in testdisplay is an i-g-t bug.
(In reply to Ander Conselvan de Oliveira from comment #6) > > And is this a kernel bug or igt bug? > > Well, the test indicates Damien's patch work as intended. Using an ARGB8888 > fb for the primary plane in platforms other than SKL+ shouldn't work. > > The crash in testdisplay is an i-g-t bug. Hi Ander, I need to double check here, does it mean: 1. "testdisplay -d 32" should report "Invalid parameter" for -SKL platforms (for example, BSW, BDW, HSW etc) 2. "testdisplay -d 32" should display normally for SKL+ platforms (SKL, BXT etc). ?
(In reply to Jeff Zheng from comment #7) > (In reply to Ander Conselvan de Oliveira from comment #6) > > > > And is this a kernel bug or igt bug? > > > > Well, the test indicates Damien's patch work as intended. Using an ARGB8888 > > fb for the primary plane in platforms other than SKL+ shouldn't work. > > > > The crash in testdisplay is an i-g-t bug. > > Hi Ander, I need to double check here, does it mean: > > 1. "testdisplay -d 32" should report "Invalid parameter" for -SKL platforms > (for example, BSW, BDW, HSW etc) > 2. "testdisplay -d 32" should display normally for SKL+ platforms (SKL, BXT > etc). > > ? Yes to both.
With the igt_remove_fb fixed there's nothing left to do here.
Keeping this open until someone pushes the fix for i-g-t. (I don't have access)
The same issue also happened on the BDW with the drm-intel-testing-2015-05-22 kernel,wish the fix for i-g-t as soon as pssible.
Assuming fixed now.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.